Glossary
Taxonomy
Also called: information architecture, folder structure
A taxonomy is the structure by which documents are organised and found — folders, categories, document types and the metadata that describes them. Its quality determines whether people can find things, and its maintainability determines whether they still can in three years.
Taxonomy explained
Folders, metadata, or both
The long-running argument in document management. Folders force a document into one location when it may legitimately belong in several; metadata lets the same contract appear under the client, the type and the renewal quarter without duplication.
Metadata-first is the better architecture when the model is maintained. Folders are more robust when it is not, and they are immediately legible to someone who joined last week. Most working systems use folders as the backbone and metadata for filtering, reporting and retention.
The test that matters
Will this structure survive the person who designed it leaving? A taxonomy maintained by one enthusiast degrades quietly once they move on: value lists accumulate near-duplicates, new document types get filed against approximate categories, and retrieval quality drops slowly enough that nobody raises it as a problem.
Design for the organisation you have, not the one that would use your model correctly.
Practical guidance
Mirror how work is actually organised — by matter, by asset, by consignment, by counterparty — rather than by department, because departments reorganise and the work does not. Keep the depth shallow; a hierarchy more than four levels deep is usually encoding metadata as folders. And name a single owner with authority to say no to additions.
FAQ
Taxonomy: common questions
How deep should a folder hierarchy be?
Rarely more than four levels. Beyond that you are usually encoding as folders what should be metadata, and each additional level multiplies the number of places a document could plausibly be filed.
Should we reorganise before or after migrating?
Migrate the existing structure, then reshape it once you can see how it is used. Changing structure and system at once makes it impossible to tell which change caused a problem.
Related terms
- Check-in / check-outCheck-out locks a document for editing by one person and shows others who holds it; check-in releases the lock and creates a new version.
- Common data environmentA common data environment is the agreed single source of information for a project, through which all information is collected, managed and shared.
- Contract lifecycle managementContract lifecycle management covers the whole life of an agreement: request, drafting, negotiation, approval, signature, storage, obligation tracking and renewal.
- Document management systemA document management system is software that captures, stores, versions, secures and retrieves an organisation's documents from one repository.
- Document migrationDocument migration is the movement of documents, their metadata and their version history from one system to another.
- Document numberingDocument numbering assigns each document a unique identifier, generated by the system according to rules configured per document type.
Last reviewed: August 28, 2026. Browse the full glossary.